INDEX

Explanations

words related to intentions, actions, and promises

oai_token-act-pair · gpt-3.5-turbo

terms related to activities, intentions, actions, and promises

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 11-res-jb

Configuration

jbloom/GPT2-Small-SAEs-Reformatted/blocks.11.hook_resid_pre

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

24,576

Data Type

torch.float32

Hook Point

blocks.11.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

asus

-0.65

 Bucks

-0.60

 Coat

-0.59

bilt

-0.59

 Bulldogs

-0.59

 Rutherford

-0.57

ossal

-0.56

helm

-0.55

 Falcons

-0.55

wise

-0.53

POSITIVE LOGITS

etting

1.17

heet

1.17

mith

1.17

cale

1.08

creen

1.07

etter

1.05

chool

1.05

paces

1.03

pring

1.03

et

1.01

Activations Density 0.361%

words related to intentions, actions, and promises

terms related to activities, intentions, actions, and promises

No Comments

No Known Activations